评估增强型与标准三维训练模型在腹腔镜技能习得中的有效性：一项新手医学生的随机对照试验

本研究旨在比较使用增强型三维（3D）训练模型与标准3D LapPass模型在新手医学生中腹腔镜技能习得的优劣。我们假设采用低保真度、具有更高维度与空间复杂度的3D增强模型，能加快适应单眼深度感知线索，从而提升表现。采用前瞻性单中心两臂随机对照试验，将参与者随机分为3D增强组和标准3D组（比例1:1，块大小2和4），完成两项标准化腹腔镜任务：1）棍子抓取与操作，2）腔内缝合。通过全球腹腔镜技能评估（GOALS）工具、完成时间和错误数进行表现评估，在基线、中期和训练后对每项任务进行测评。研究结束后通过问卷调查，评估参与者的信心和表现（注册编号：NCT06184854）。英国伦敦皇后玛丽大学巴茨癌症研究所共招募65名医学生，全部为腹腔镜手术新手，完成研究的有62人（其中3人退出）。两组在人口学特征上无差异（p>0.05）。两组在两项任务中训练后表现均显著优于基线（组内比较，p<0.017）。在组间比较中，3D增强组在GOALS评分和深度感知方面的改善显著优于标准组（p<0.05），时间改善和错误减少趋势亦优于增强组但未达统计学意义。问卷中，增强组在缝合任务中的信心和表现感知显著优于标准组（p<0.05）。结论：使用增强型3D训练模型能显著提升新手在两项腹腔镜任务中的技能水平，建议将其纳入常规培训课程以增强学习效果。未来应关注其长期效果及以能力为导向的评估指标。

The aim of the study was to determine if laparoscopic skills acquisition will be superior with the use of 3D-enhanced training models with increased dimensionality and spatial complexity compared to standard 3D LapPass training models in novice medical students. We hypothesized that adopting low-fidelity, 3D-enhanced training models in the 2D visual modality may expedite adaptation to monocular cues related to depth perception, enhancing performance.A prospective, single-center, 2-arm, randomized controlled trial with participants randomized equally into either 3D-enhanced or standard 3D group (allocation ratio 1:1, block sizes 2 and 4) performing 2 standardized laparoscopic tasks: (1) polo grasping and manipulation and (2) intracorporeal suturing. Performance was assessed using the Global Operative Assessment of Laparoscopic Skills (GOALS) tool, completion time, and number of errors, with evaluations at baseline, interim, and post-training for each task. A post-study survey measured participant-perceived confidence and performance (ID: NCT06184854).Barts Cancer Institute, Queen Mary University of London, UK.Sixty-five medical students, all novices in laparoscopic surgery, were enrolled, with 3 dropouts.Thirty-one participants in each group completed the study with homogeneity regarding demographic data (p > 0.05). Both groups had significantly superior performance post-training compared to baseline in the 2 tasks (intragroup comparison, p < 0.017). In the intergroup comparison, the 3D-enhanced group had statistically significantly greater improvements in the GOALS score parameter and depth perception domain across all comparisons for both tasks (p < 0.05). Time improvement and error reduction also favored the 3D-enhanced group, however, without reaching statistical significance. In the post-study survey, the 3D-enhanced group reported significantly greater perceived post-training confidence and performance in the suturing task alone (p < 0.05).The use of 3D-enhanced training models resulted in superior laparoscopic skills acquisition for novices in 2 tasks compared to standard 3D LapPass models. Therefore, incorporating 3D-enhanced models into laparoscopic training curricula alongside standard 3D LapPass models can supplement learning. Further work should focus on long-term efficacy with a competency-based endpoint.